EXPLORATION OF SALES AND FINANCE FOR THE FISCAL YEAR JANUARY, 2019 TO JANUARY, 2020

By;

RICHARD DANQUAH

31st January, 2020

A REPORT PRESENTED TO THE FINANCIAL TEAM AND MARKETING MANAGEMENT ON SALES TRENDS AND SALES RETURNS FOR THE FISCAL YEAR’S FINANCIAL SALES STATEMENT.

1 Introduction

This Report explores the possibility of using Python Programming as tools to automatically process data gathered from financial activities and extract relevant information to achieve the goals of the Marketing and Finance Team. The objective is to perform an automatic process of data using artificial intelligence techniques, in particular the application of machine learning algorithms. The final goal is to manage the extraction of quantitative data with relevant information to inform manager about the current competitiveness of our products and services.

1.1 Staement Of the Report

This report is intended to inform managment on the performance of the Marketing managers on how their marketing strategies were able to reach both loyal and impulse customers and improved Sales Value. This finance report will communicates the Firm's crucial financial information and competitiveness and covers a specified period through daily, weekly, and monthly financial reports. This will be our tools that you can apply to increase internal business performance as compared to privious years. The data on the finance report is also an effective means of remaining updated with any significant progress and changes in the status of our finances and guid as to measure our financial results, cash flow, and financial position on Monetary terms.

1.2 Objectives Of The Report

-- To Understand our current customer base in all States

-- To understand our Financial Position on how much Sales was declared 

-- To Understand Buying trends base on Location and Day of The Week

-- To understand with of the Month or Day of the Week is suitable for promotion

-- To understand how many items were sold through out various Days of the Week and Months.

-- To understand most items sold and how much sales was made from each item sold in the Fiscal year. 

2.0 Data Preprocessing

Importing Important Python Libraries

Loading Data Into Pandas DataFrame

We call our Data Frame to print out our data in our Database or in our repositry and print the first five (5) rows of our Data frame.

We check in our data set if there are unusual entries in the data set, if there are any, python will automatically compute all our unusual entries and tell us which of our columns and rows contains these errors or entries. Where in the table indicates True means an error entry and where False means vice versa.

We use Our Pandas Library to process the data to give us information on our data set of which belongs to which data type in our Data Frame. From the table we can see our columns corresponds to various Data types in a Data Frame.

We call a new Variable and redefine Our Colums respectively

We then Check our columns we priviously defined in our Data Columns to a structured column names and print out our first five row of our dataset.

After examine our data set and having a second look at the error entries, we then call our Python Pandas magic library to collate and remove all error entries in our dataset. From the table above We can see a sample of the Order ID, Product purchased, Quantity Ordered, Price of the product, the Order date of Purchase and Addresses of customer.

Checking different time stamp entries in our Data

We then Count how many time a product was bought by different customers in different times. This is not the actual total sum of Order quantity purchased by our customers. The dataFrame above shows that USB-C Charging Cable was purchased 21903 different times, followed by Lightning Charging Cable 21658, AAA Batteries (4-pack) 20641, AA Batteries (4-pack) 20577, Wired Headphones 18882 and as follow in order on the DataFrame.

The table above gives us a fair Idea of how many time or quantity of items bought. Number of quantity of items a customer, From our Data output we can see that 168552 customers bough 1 items at a time, 13324 customers bought 2 items at a time, 2920 customers bought 3 items at the same time of purchase. Again, 806 customers boughr 4 items at a time, 236 customers also bought 5 items at a single time of purchase, 80 customers bought 6 items bought at the same time.Lastly, 24, customer, 5 customer and 3 customers also bought 7, 8, and 9 Items in a single time of purchasing respectivey. Statistically, We have a fair idea that, Most of our customers are likely to buy 1 or 2 or 3 items at a time. This will give managers and decision makers to plan of packaging efficiency and optimize packaging cost.

Data Munging and Features Extraction

We are converting all our our columns such as QuantityOrdered, PriceEach, OrderID, Product, PurchaseAddress to their respective Data type in Pyhton for further Analytics to be carried out. Lastly we use the TimeDate Stamp to convert our dates into proper Python date type.

This our first descriptive Analysis in depth, In Data Analytics, one of the most important step is have astatistical idea of the numerical data types in Dataset. The Count of the summary of all data entries, Mean, that is the avarage score, Standard deviation, which is a deviation or how far apart a deviation in a dataset is from the total mean. The minimum value and the Maximum value in the dataset, the first quatile, the second quatile known as inter quatile and the thrid quatile. This is mostly visualized using the whisker box plot to check for skewness in our data set or for outliers. This is mostly and also important when drawing our normal distribution curve with the numerical values in a Data

Munging Date In Python

Munging Sales Feature From Price and Quantity

We call upon a Magic Pandas library in Python to extract the month from the actual Date and create a new column call "Month", Lastly, we again use similar process to extract our total sales per day from our dataset. This process is one of the most significant steps in this Analysis since the Data we extracted from the Database did not contain out total sales made in a day, but we are confident and with competency in Python that we can in any way computed our total sales and further create a corresponding column to hold our total sales column in our dataset.

We count the Number of Items sold in each month from Jenuary to December.

Extracting With Lamda, City From Purchase Address

We call upon a Magic Pandas library in Python to extract the Various Cities from the actual Purchasing Addresses and create a new column call "City", Lastly, we again use similar process is aslo another significant steps in this Analysis since the Data we extracted from the Database did not contain Cities but rather Addresses, but we are again confident and with competency in Python that we can extract our individual Cities Name and abbriviation from the Purchase Address to create a corresponding column to hold our individual Cities in our DataFrame.

Converting Strings, Floats and Integers

From the above codes, since we have created a new Columns into our Dataset, We need to convert all columns in their respective data type.

Munging Days From Dates With Pandas Series

We use our Magic Pandas library in Python to extract the Days of the Week from the Date stamp and create a new column call "Daysinweek" which is aslo another significant step in our Analysis since the Data we extracted from the Database did not contain the Days of the Week but rather Date of the day, with my competency in Python I will extract the Days of the Week from the Date stamp to create a corresponding column to hold our "Daysinweek" in our DataFrame.

This is the last step in Our Initail Data Preprocessing, I checked all coresponding Columns if they are attributed to their various respective Data type in other to be able to proceed to Indepth Analysis.

Total sales For the Fiscal Year

A descriptive summury of numerical values in our Dataset.

The total Sum of Sales VodaFone made in 2019. From the above computation, VodaFone made a sum of sales $ 34492035.97 in mobile accessaries and electronics.

Total Qauntity solds For the Fiscal Year

The total sales of item quantity sold was 209079 items in 2019. This gives us an Idea of how many Items were sold in 2019 so as to be able to compare with subsequent and previous years to be able to determine our competitiveness in the current market.

Computing Sales and Quantity Sold by Cities

Sum of total in quantity order of Items purchased in each City and the total Sum of sales by Various Cities.

In reference to our computation, San Francisco(CA) made to top Quantity order of items of 50239 Items and total sum of sales of $ 8.262204e+06, followed by Los Angeles(CA) with a Quantity order of items of 33289 items and a total sum of sales of $ 5.452571e+06. New York City(NY) also made a Quantity order of items of 27932 Items and a total sum of sales of $ 4.664317e+06. Boston(MA) made a total sale of Item Quantity of 22528 and a total sum of Sales of $ 3.661642e+06. Dallas(TX) also contributed 16730 of count of quantity order and a total sum of Sales of $ 2.767975e+06. Again, Atlanta(GA) made a total sum of order quantity of 16602 Items and a sum of Sales of $ 2.795499e+06, Seattle(WA) with a count on purchase of quantity order of 16553 and a total sum of Sales of $ 2.747755e+06 and Portland(OR) with a quantity order of 11303 items and a Sum of sales of $ 1.870732e+06. Austin(TX) and Portland(ME) also contributed 11153 and 2750 order of quantity and $ 1.819582e+06 and $ 4.497583e+05 sum of Sales in our Market share respectively. This Analysis tells us more about which city made the most purchase and give managers and idea of planning on segmentation, pricing, discounting and promotion in order to build upon our market share and compete well in the competitive market.

Item Quantity Of Order By Each City Within Our Segment

In reference to quantity order of Items, San Francisco(CA) made to top Quantity order of items of 50239 Items followed by followed by Los Angeles(CA) with a Quantity order of items of 33289 items, New York City(NY) also made a Quantity order of items of 27932 Items and Boston(MA) made a total sale of Item Quantity of 22528 Items. Dallas(TX) also contributed 16730 of count of quantity order whilst Atlanta(GA) also followed with a total sum of order quantity of 16602 Items. Again, Seattle(WA) with a count on purchase of quantity order of 16553 Items and Portland(OR) with a quantity order of 11303 items. Last, Austin(TX) and Portland(ME) also contributed 11153 and 2750 order of quantity in our Market share respectively. In reference to this Analysis, we have an idea which part of these cities is our strong segment to be able to make more effort in order to sustain or improve upon our competitiveness. Servearl steps could be taken to be able to make sure we do not loss our market value but rather define a Decision Surport Systems (DSS) based on Research and Analysis at why each Cities with our Segment contrinuted to their corresponding item order in able.

Ploting Sales By Cities

Plotly, a Powerful Python Library usedd to plot intereactive plots of 2 anad 3 Dimentions in Data Science. A Line Plotly plot to show a Visual of Total Sales by Cities in 2019. Line Plot with go.Scatter If Plotly Express does not provide a good starting point, it is possible to use the more generic go. Scatter class from plotly graph objects. Whereas plotly.express has two functions scatter and line, go. Scatter can be used both for plotting points (makers) or lines, depending on the value of mode.

Plotting Quantity Sold By Cities

Plotly, a Powerful Python Library usedd to plot intereactive plots of 3,4, 5 Dimentions in Data Science. A Bar plot to show a Visual of Total Quantity order by Cities in 2019.

3 Dimensional Plot For Total Quantity and Sale By Cities

Plotly, a popular Python Library used for building intereactive plots of 2 and 3 Dimentional Visualization in Data Science. A 3 Dimensional Line Plotly plot to show a Visual of Total Sales by Cities in 2019. Plotly Express is an easy to use, high level interface for Plotly library, which operates on a variety of type of Data and produces easy to style figure with px.line. Each data point is represented as a vertex which location is given by the x and y column of a polyline mark in 2D space.

Ploting Sales By Cities

MatplotLib and SK Learn, another popular and most widely used Python Visualization and Modeling Machien Learning in Data Science. The Plot above shows the distribution of Total Sum of Sales in each Segment or City.

Computing Sales and Quantity Sold By City and Month

Sales made by each City is vital in this Analysis, It gave us full adea of how much each City contributed to total sales per Month. The summury is very usefull in identifying how much revenue was returned from all monthly purchasese. The Analysis also creat a window for Maximization and planning both short and long term Forcasting. In the Management of finance on Budgeting which is a keey financial plan of business attempting to forcast a number of months ahead, it is very important to know how much in each month, each City responded to our products and the revenue we made in each month by each City or Market segment. From our Analysis, in the month of January, customers at San Francisco(CA) made the highest purchase of 2661 and a total sum amount of $ 435588.33, followed by Los Angeles (CA) with a purchases of 1716 Items and a total sum amount of $ 288601.90. Boston(MA) also contributed 1174 of tottal items sold and a total sum amount of $ 201088.49 and Atlanta(GA), 876 Items and a total sum amount $ 149159.54. Again, Seattle(WA) also purchased 810 Items and contributed to $ 140692.05 of total sales made in January 2019, the City of Portland(OR) purchased 555 Items anda total sum amount $ 92276.76. Austin(TX) and Portland(ME) with 582 and 138 Items contributiing to 88087.06 and 22708.80 respectively.

In the Month of February San Francisco(CA) with the highest purchases of 3243 Items contributing sum amount of $ 547072.34, Los Angeles(CA) being the second higest of 2160 Items and sum amount of $ 342061.66, followed by New York City(NY) with 1805 Items and sum amount of $ 305372.26, Boston(MA) third of competotive order of 1400 Items and also contributing $ 214808.36 and Dallas(TX) with 1112 Item and a total sum of sales $ 186667.99. Again, Atlanta(GA) petronised most of our products with an order of 1042 items and contributed to $ 176470.30 of our Total sum of sales whilst Seattle(WA) was next on competitive order of 1066 Items and sum of sales of $ 171330.25. Portland(OR) also compete with 728 Items and sum of sales of $ 119606.37. Last, Austin(TX) and Portland(ME) with quatity order of Items of 685 and 208, sum of sales of $ 108787.40 and $ 29845.49 respectively.

In the Month of March, 2019, San Francisco(CA) still toped in ranking order of purchases of 4025 Items and sum of sales of $ 693726.96, Los Angeles(CA) with 2755 order of purchases and sum of sales of $ 429929.37, followed by New York City(NY) with 2211 Items of order and sum of Sales of $ 367262.20, Boston(MA) also was on the log with 1881 order of Items and sum of sales of $ 301023.81. Atlanta(GA) with 1384 Items anad sales of $ 231905.38 whilst Dallas(TX) made a total purcgase of 1349 items and Sales returns of $ 222376.68. Seattle(WA) with 1331 Items of order and sum of $ 219118.70. Portland(OR), with 903 Items and contributed 156691.72 of our total sales in march. Last, Austin(TX) with 952 order of items and sales returns of $ 154549.27 and Portland(ME) 214 Items and sales of $ 30516.29.

In April, 2019, San Francisco(CA) again leading of order of Items of 4975 and $ 812426.19 of sales, Los Angeles(CA) with 3398 Items and $ 550264.02 of Sales, New York City(NY) also with 2746 items and Sales of $ 449447.75. Boston(MA) 2189 Items and sum of Sales of $ 353807.11, Atlanta(GA) also made a total order of 1634 Items contributing to $ 284448.91 of our total sales in April. Customers in Seattle(WA) with 1609 items and $ 276254.07, Dallas(TX) also on the order with 1518 Items and $ 251360.48 of sales. Portland(OR) with 1132 Items and $ 197441.63 of sales returns whilst Austin(TX) was next of the log with 1092 Items and $ 172683.59 of sum of sales and Portland(ME) with the least orders of 265 Items and $ 42536.49 of total sales.

In May, San Francisco(CA) obviously lead the log again with the highest purchase of 443 Items and sum of sales sales $ 776679.49, Los Angeles(CA) second highest with 2885 items and $ 499689.21 of sales, New York City(NY) with 2555 Items and $ 436126.40 of sales, Boston(MA) with 1991 Items and sum of sales of $ 328803.65 and Dallas(TX), 1616 Items on order and sum od sales of $268456.49. Followed by Atlanta(GA) with 1502 Items and cntributing to $ 238853.99 of sales returns, Seattle(WA) with a purchase order of 1405 Items and $ 211654.29 of sales of order. Portland(OR) with 1068 items and $ 173729.25 as sales by total order. Last, Austin(TX) with 961 items and $ 160635.22 of sales and Portland(ME), 241 items and $ 57978.76 of sales returns on order of purchase in the month of May, 2019.

In June 2019, San Francisco(CA) contributed the highst order 3687 Items and $613173.48, Los Angeles(CA) was the second highest with an order of 2452 Items and $ 451531.93 or sum of sales of order, New York City(NY) with 1999 Items and sum of sales of $ 324148.58, Boston(MA) next with 1614 items quantity order and $ 254747.89 of sum of sales. Atlanta(GA) also contributed heavily on quatity border of purchases of 1231 Items summing up to $ 219816.47 of Sales, Seattle(WA) with 1198 Items of quantity order and a sum of sales of $ 213853.52, whilst Dallas(TX) also made a total purchase of 1188 Items and $ 186885.75 of sum of Sales. Moreover, Austin(TX) with 803 items and $ 144057.29 of sales, Portland(OR) with quatity order of 850 items and $ 139562.02 of Sales and Portland(ME) last on the log order with 231 Items and $ 30025.33 of sum of sales in June only.

Our Analysis shows that in July, 2019, San Francisco(CA) with a quantity order of 3994 Items and $ 642881.76 of sales, Los Angeles(CA) again as second highest with 2501 quatity order and $ 394334.64 of sales, New York City(NY) with 2168 Items on order summing up to $ 355716.10 of sales, Boston(MA) follows with 1772 items on order and $ 291497.14 on sales. Seattle(WA) with 1265 on orders and $ 212513.92 on sales, Dallas(TX) also with 1236 on quatity order and $ 212325.17 on sum of Sales, Atlanta(GA) with 1259 Items on purchase and $ 211766.47 on sum of sales, Austin(TX) making a total order of 850 Items and $ 150324.93 as sum of sales on quatity order. Last, Portland(OR) with 809 items on order and $ 143994.49 as sum of sales and Portland(ME) with 218 items and $ 32421.14 as sum of sales on July only.

In the Month of August, San Francisco(CA) still leading in total purchase with 3191 of quatity order and $ 538778.10 of sum of sales, Los Angeles(CA) stiil second highset in oreder quatity of 2076 items and $ 345893.50 of sales, New York City(NY) with 1809 quatity order and $ 303183.46 of sales, Boston(MA) also with 1408 items on order and $ 239275.26 as sum of sales. Seattle(WA) made a total order 1147 items and $ 189715.09 as sum of sales, Dallas(TX) with 1102 Items and sum of sales of $ 179763.46, Atlanta(GA) with 1053 Items and $ 169267.66 of sales, Austin(TX) with 735 Items and $ 125713.61 of total sales. Portland(OR) with 745 qauntity of order and $ 116881.14 of sum of sales and last, Portland(ME) with 182 of quantity order of items and a sum of sales of $ 35996.60 for ther month of August only.

The distrubition table again shows that in the month of September 2019, San Francisco(CA) as usual leading on quatity of order with 3083 Items and $ 463595.72 as sum of sales, followed by Los Angeles(CA) with 2123 items of order and $ 354075.69 as sales returns on order, New York City(NY) also with quantity order of 1708 Items and $ 300563.83 as sum of sales of order. Boston(MA) made a total quantity order of 1456 items and $ 248408.73 as sales, Atlanta(GA) with 1035 Itmes and a sum of sales of $ 171278.89, Dallas(TX) as one of our competitive market segments also with 1053 of total orders and a sum of sales on order of $ 164212.86. Again, Seattle(WA) also with 1059 iteme on order and $ 156369.27 of sum of sales and Austin(TX) with 763 items and a sales of $ 106483.70. Last, Portland(OR) and Portland(ME) with order quntity of 668 and 161, sum of sales of $ 103811.88 and $ 28759.56 respectively for the monthe of September only.

In October, San Francisco(CA) lead the table with 5310 items on order and summing up to $ 66700.98, Los Angeles(CA) second highest with 3725 on order of quntity and $ 612453.25 of total sum of sales, New York City(NY) with 2951 of item order and sum of sales of $ 486954.41, Boston(MA) also contributed 2444 items on total purchase and $ 367036.39 of sum of sales. Dallas(TX) with 1813 items and sum of sales of $ 323135.60, Seattle(WA) with 1877 items and $ 316856.26 as sales return, Atlanta(GA) also with order total of 1793 items and $ 306293.01 as sales rreturn on purchases, Austin(TX) with 1235 items and $ 203196.12 of sales. Last, Portland(OR) madea total quatity of purchase of 1274 items and and $ 201,778.34 sales on order and Portland(ME) with 281 items of order and sum of sales of $ 52,322.52 for the Monthe oc October only.

In November, San Francisco(CA) with the highest purchases of 4840 Items contributing sum amount of $ 764979.29, Los Angeles(CA) being the second higest of 3120 Items and sum amount of $ 499690.79, followed by New York City(NY) with 2651 Items and sum amount of $ 428180.32, Boston(MA) third of competotive order of 2173 Items and also contributing $ 351546.02 and Atlanta(GA) with 1568 Item and a total sum of sales $ 275338.70. Again, Seattle(WA) petronised most of our products with an order of 1583 items and contributed to $ 252080.13 of our Total sum of sales whilst Dallas(TX) was next on competitive order of 1562 Items and sum of sales of $ 248609.56. Portland(OR) also compete with 1065 Items and sum of sales of $ 173,210.70. Last, Austin(TX) and Portland(ME) with quatity order of Items of 977 and 259, sum of sales of $ 1712,86.47 and $ 34,681.22 respectively.

In December, this is the moth that shows our booming market share more than other moth in the year of 2019 comparing all quantity of order and sum of sales by each City. The distrubition table again shows that in the month of December 2019, San Francisco(CA) as usual leading on quatity of order with 6787 Items and $ 1,106,601.27 as sum of sales, followed by Los Angeles(CA) with 4378 items of order and $ 684,044.84 as sales returns on order, New York City(NY) also with quantity order of 3816 Items and $ 646,770.83 as sum of sales of order. Boston(MA) made a total quantity order of 3026 items and $ 509,599.16 as sales, Seattle(WA) with 2203 Itmes and a sum of sales of $ 387,317.93, Dallas(TX) as one of our competitive market segments also with 2303 of total orders and a sum of sales on order of $ 380,718.85. Again, Atlanta(GA) also with 2225 iteme on order and $ 360,899.26 of sum of sales and Portland(OR) with 1506 items and a sales of $ 251,748.04. Last, Austin(TX) and Portland(ME) with order quntity of 1518 and 352, sum of sales of $ 233,777.09 and $ 51,966.07 respectively for the monthe of December only.

Analysing and Comparing Sales In Each Month

A "Stacked Bar plot" with an interactive Plotly Library in Python to Visualized our distribution of total Sales anad quantity order by each City and each month. From a statistics point of view and a mere observation December was our booming market and sales retunrs and this could be a factor of Christmas which likely, every producer is likely to earn more income. Top manager can use this period a source of comperitive advantage so as to target customers of specific months and its effects on our products ad service so as to creat, gain and maintain competitive advantage by creating a unique and a core resiources and compitencies in the form of promotions and advertisments.

A bubble plot with an interactive Plotly Library in Python to Visualized our distribution of total Sales anad quantity order by each City and each month. From a statistics point of view and a mere observation December was our booming market and sales retunrs and this could be a factor of Christmas which likely, every producer is likely to earn more income. Top manager can use this period a source of comperitive advantage so as to target customers of specific months and its effects on our products ad service so as to creat, gain and maintain competitive advantage by creating a unique and a core resiources and compitencies in the form of promotions and advertisments.

Analysing and Ploting Quantity Sold By Cities

Since there are different ways of Visualizing they I presentesd the Total Sales by each City into a different data Visulization form to have the best explanation and a better understanding of the bout put for our Data. Above is a Bar plot of Total Sum of sales by each City in our Market segment.

Analysing Total, Average, Maximum and Minimum Sales By Cities In Each Month

The results of the distribution table tell managers about a descriptive Statistics of how each City within our Market segment performed in each month as previous Analysis but this time we need to understand other statistics such as "Mean", that is the average sum of sales or the average sales per month. As a Company Junior Analyst, I will need to put more interpretation on this result to be able to communicate properly to manager and all stakeholders of the company. First, the Column "City" returns all Cities in our Marketing segment, Second, the column "Month" is the particular month in which sales was sum up together to arrive at an actual sum of total sales in that particular Month. Third, the Column "Sum", this is the Total summation of all Purchases that is ("Total Quantity of Order") * ("Price Each"). Again, "Mean", that is the average sum of sales or the average sales per month, the Column "Max", this is the Maximum price of an Item purchased with the month and last, "Min" which refers to the Minimum item price ordered in that particular Month. From this observation, we are certain and sure that, the minimum item price for our product purchased by our customers was $ 2.99 and the Maximum price of our products is $ 1700.00. In a Decision support System, Strategic Managers can evaluate this performance on Budgetary control, where management tool is a key element to set realistic performance target and then monitor subsequent results base on City Average and total sales per month to see whether or not our business is on the right course and to be able to provide aa definite goal and objectives that serves as benchmark to evaluate performance.

Analysing Total, Average, Count, Maximum and Minimum Sales on Items by States in Month

In the privious Analysis we were certianed and clear about how cities, Month react to sales and it beahavious towards eachother mathematically and in a corrilation matrix. In this Analysis we Analyse which specific Items were bought in a particular Month. In a destribution of Products accorss various Cities, we will be confidence and able to monitors our product sales returns against privious sales. Giving more interppritation to Our above Analysis table above, In the City column is the Particular City, the Month is the exact month of sales, the Product column also tell us about which products werte bought in that particulr month. The Sum is a Pandas sub-index of sales which is the Toatal Sales returns by a specific item total sales in that actual month, the mean is the "Avarage sale of that item only" in that particular month in Analysis. The Count Column show how many different times that particular Item was bought (Not the "Total Quantity Order" of the Item) in the case of 20in Monitor in January at Atlanta(GA) was bought 12 different times by either same or different customer on different occasion with a Maximun returns of $ 220.00 meaning non of our customers bought this item more that two(2) on a sale at a go and a Minimum of $ 110.00, which explains the actual price of product. The same interpritaion are as follows to the last City on our Analysis table to the last month. As a Junior Data Analyst and a Business Analyst, will be able communicate with figures aiding Managers to understand which Item returns a high profit at a prticulr Month and by a Paticular City and which Item performed least. this will help to enlighting us on how to do proper planning of which Items are profitable and need to produce more than those that do not perform better. This Analysis will be valuable when palnning for a Market developmet startegy such as product Developemt and Diversification strategy.

Analysing Qunatity Sold in Months and Sorting in Ascending order

The DataFrame above shows how many Items were sold and how much Sales was made in each Month from January to December. The table was presented from the highest quantity of Order and Sales in each month. Clearly it is obvious that December is the most highest sales and quantity order followed by

In the Best Month Analysis, the idea is the give a full accounts and details of how much Total Sales was returned in different month. The table was presented from January (as Month 1) to December (as Month 12),the actual total quantity of Order and Total sum of Sales in each month. In January we made a Total sum of sales of $ 1,822,256.73 and a total quantity order of 10903 Items, In February, we made a Total sum of sales of $ 2202022.42 and a total quantity order of 13449 Items, in March, our total sum of sales was $ 2807100.38 and total quantity order of 17005 items. Again, in the Month of April, $ 3390670.24 was our actual total sales and a quantity order of 20558 Items whilst In May, we made total sales of $ 3152606.75 and 18667 order of quantity, in June, our customers were still responding to our services and product and we were able accumulate a total quantity order of 15253 items and a Total sum of sales of $ 3390670.24. In the Month of July, quantity Order apricated up to 16072 ordered items and a Total sum of sales of $ 2647775.76 against June. In August, we recorded a little low quantity order of 13448 items and $ 2244467.88 of total sum sales as compared to previous months. Moreover, in September, total quantity order and Sales depreciate again to 13109 quantity order and $ 2097560.13 as sum of sales for September, In November there was little improvement on quantity order and Total sum of sales to 22703 Items on order quantity and $ 3736726.88 as total sum of sales whilst in November, there was again a depreciation to 19798 items and $ 3199603.20 on sales returns as against October. last In December, this month was where we performed better in quantity order and total sales. In December we recorded 28114 items and a total sum of sales of $ 4613443.34. This could be attributed to the fact that Christmas might influence our high production.

A highly interactive Plotly 3D Plot show how moths performed against quantity order and Toatal sum of sales in 2019.
A highly interactive 3 Dimension Bubble plot show how moths performed against quantity order and Toatal sum of sales in 2019.
A MatploLib Line Chart of Total sum of sales in Month, from Month (1), January to Month (12), December. The Visyalizatio shows various tickers of months in a red tick which indicates a presence of a prticular month.

Analysing Sale and Quntity By Days Of the week

In this Analysis, we exploit an Idea behind a specific day of the week and how much sales we earn, this Analysis is mostly important we want to identify patterns in a Data that tells manager when to we should expect most customers buying from our shops. this could be attributed to so many factors. From our reliable Analysis, most of our customer buys electronics from our shop on Tuesdays which recorded 30724 items and $ 5,087,957 total sum of sales from the first Tuesday of the year to the last Tuesday of 2019, followed by Wednesdays which recorded 29799 items and a total sales of $ 4,988822, from the first Wednesday to last Wednesday in 2019, Sunday also was another booming market sales of $ 4,932,170 and a total quantity order of 29778 items from the first Sunday to the last Sunday in 2019. Again, in all Saturdays in the year, we recorded 29769 total quantity of items and $ 4,904,357 of total sum of sales, Mondays was a challenging day in sales and quantity order but our sales representative’s mange to make a total sum of sales of $ 4,883,327 and 29899 total quantity order which was not extremely bad. On all Fridays our sales were not encouraging but yet we manage to settle with our provisional and expected minimum quantity order of 29448 Items and a total quantity of order of $ 4,855,938. Last, on Thursdays, we recorded our least sales and quantity order even thou Mathematically, we were still able to maximize our daily sales target. We were initially thinking most of our customers may buy much from us on weekends hypothetically, but our thoughts were wrong.
A Plotly Bar plot Showing each Day of the Week and how much Total we earned from the total sales from first to Last Day of the Week.
A MatplotLib Bar plot Showing each Day of the Week and how much Total we earned from the total sales from first to Last Day of the Week.

Analysing Sales and Quntity Sold By Specific Items

Identifying which product or products sells more in our sales Analysis is very crusial in Accounting concept. In most cases it give us a high level knoledge of as to whicuh products performed better, to able to understand the preference of our customers and future demands of such product or item.

Identifying which product or products sells more in our sales Analysis is very crucial in Accounting concept. In this Analysis we consider the item with the highest order quantity, our focus is on total number of quantity of items by each product. In most cases it give us a high level knowledge of as to which products performed better, to able to understand the preference of our customers and future demands of such product or item. From the Analysis table, AAA Batteries (4-pack) was the most sell out items in 2019 with a quantity order of 31017 items and a Total sum of Sales of $ 92,740.83 and at $ 2.99 each item price, followed by AA Batteries (4-pack) which also sold out 27635 total sum of quantity order and $ 106,118.40 sum of sales at $ 3.84 per item price, USB-C Charging Cable also followed with total quantity order of 23975 items and a total sum of sales of $ 286501.25 at item price at $ 11.95, Lightning Charging Cable recorder total quantity order of 23217 items and a total sum of sales of $ 347094.15 at a unit price of $ 14.95, Wired Headphones also contributed 20557 quantity order and at total sales of $ 246478.43 at a unit rpice of $ 11.99. Apple Air pods Headphones with quantity order of 15661 orders and a total sales of $ 2349150.00 at a unite price of $ 150.00 whilst Bose SoundSport Headphones with a quantity order of 13457 orders and a total sales of $1345565.43 at a unit price of $ 99.99, 27in FHD Monitor made a total order of 7550 and a total sum of sales of $ 1132424.50 at a unit price of $ 149.99. Moreover, 34in Ultrawide Monitor with a quantity order of 6199 and a total sales of $ 2355558.01 at a unit price of $ 379.99 as compared to Google Phone with a quantity order of 5532 purchases and a total sales of $ 3319200.00 at a unit price of $ 600.00. Flat screen TV followed on the table with 4819 quantity order and a total sales of $ 1445700.00 and a Unit price of $ 300.00, whilst Macbook Pro Laptops were ordered a total quantity of 4728 and $ 8037600.00 at a unit price of $ 1700.00. Furthermore, ThinkPad Laptops were ordered 4130 in quantity order and $ 4129958.70 sum of sales at a unit price of $ 999.99. 20in Monitor recorded 4129 in quantity order and $ 454148.71 as sum of sales and at a unit price of $ 109.99. Vareebadd Phone was one of the few products that recorded low quantity order of 2068 orders and but made high sum of sales of $ 827200.00 at a unit price of $ 400.00. Last, LG Washing Machine was on low quantity order at 666 orders and at a total sales of $ 399600.00 at a unit price of $ 600.00 whilst LG Dryer was the least order quantity at 646 units and a total sum of sales of $ 387600.00 at a unit price of $ 600

A MatplotLib Visualization of Total Sum Of Order Qunatity Of Products.

Statistics

Now We observe our Nimimum unit of item price and the Maximum Unite price of our product. The Analysis tell us that, our Minimum unit price is $ 2.99 which is AAA Batteries (4-pack) and our maximum unit item price is $ 1,700.00 which is Macbook Pro Laptop. Now mangers have an indepth idea of a Sales analysis and the perfromace of our prouct on the competitive market.

Box Plot and Whiskers

Tukey's box plots (Robert McGill, John W. Tukey and Wayne A. Larsen: "The American Statis-tician" Vol. 32, No. 1, Feb., 1978, pp. 12-16)are statistical plots that provide useful informa-tion about the data distribution such as skewness. However, bar plots with error bars are still the common standard in most scientific literature, and thus, the interpretation of box plots can be challenging for the unfamiliar reader. The figure below illustrates the different visual features of a box plot. Violin plots are closely related to box plots but add useful information such as the distribution of the sample data (density trace).

Correlation Using Python Pandas, Association and Causation

Data Correlation Method

Data correlation is a method that helps to predict one attribute from another attribute and is used as a basic quantity in many modeling techniques. If one feature increases, the correlation will be positive, so the other feature increases as well and negative if one feature increases there will be a reduction in another. If there is no relation between any two attributes then it is said to be no correlation. If there is a linear relationship between the constant variables then the Pearson correlation coefficient is used. If there is a non-linear relation between the constant variables then the Spearman correlation coefficient is used. Since the considered data set is linear so the Pearson correlation coefficient is used for the selection of features in this study. This correlation for all the attributes is shown in figure 3.4. To improve the efficiency of the Machine Learning model, the attributes that have negative correlations were removed. It is a statistic measuring the linear correlation of two variables X and Y. It has a value between +1 and 1, where 1 is a linear positive correlation, 0 is not a linear correlation and 1 is a linear negative correlation. The motivation for considering the correlation is when people know a score on one measure, they can make a prediction of another measure that is highly related to it more accurate. The more accurate the prediction, the stronger the relationship between the variables.

The Box plot show the distribution of our sum of sales of different items, we have understood that, our sale are spread apart and other product sales seems to far apart from other sum of sales of items. This might be an influence of one item being high priced than others due to thier values. For instance AAA Batteries (4-pack) which is sold at $ 2.99 might surely be far apart from ThinkPad Laptop which is sold at $ 999.99 as a unit price. In this case we have understood from the regression plot that there is an association and a predictor of "Total Sum of Sales" on product of by price Of products not quantity of purchase or order, when price of product goes up or is higher, there is a likelyhood that our Total sum of sales will sure be higher, if all things being equal we are able to sustain or improve upon our Markerting strutures and customers are likely to petronize our products. The Correlation matrix show an association or a causation between Price of Item and Sale at a correlation Coefficient of +0.9, which means there is a Very strong positive relationship, all other Vaariables show a week or no existence and influence on each variable as being a predictive variable of the other at a -0.1 Correlation Coef.

The distribution plotstell us how our Unit price of our products are setup, from the plot we can see most of our products are priced between $ 2.99 and $ 540.00

Modeling Regression With Ordinary Least Square (OLS)

Modeling Linear Regression With OLS

OLS shows a trend of Sales. This shows that sales were affected by the Particular Month. We are confident when we should expect high sales and low sales and plan how reach more customers by the Marketing team.The OLS plot gives an insight into Sales performance in different Months.

OLS shows that Quntity sold has high influen on how much sales we made in all months, this means our price will not hinders our target sales and qunatity if only we are able reach more customers.

Note: Hover Over The Plot For PREDICTED Values Against r-Square Vales and ACTUAL Values